ABSTRACT The HIV-1 envelope protein (Env) can be presented in many forms. Functional Env, recognized only by neutralizing antibodies, is a trimeric complex of gp120/gp41 heterodimers. Non-functional forms of Env include monomeric gp120, uncleaved gp160, gp41 stumps and gp120/gp41 monomers. Recent evidence suggests that non-functional forms of Env exists on the surface of live HIV-1 and may be prime antibody targets, and perhaps account for the generally poor neutralizing responses to natural infection. Our goal is to develop a virus-like particle (VLP) vaccine that is modified to refocus antibody responses against trimers. In Specific Aim 1, we will develop tools to investigate HIV+ serum antibody responses and VLP immunogenicity (Aims 2 and 3). We will clone a panel of trimers on VLP surfaces. We will make targeted mutants in gp120 and gp41 for mapping studies (Aim 2). We will develop methods to increase the production, quality and immunogenicity of VLPs (Aim 3). In Specific Aim 2, we will probe the topology of VLP Env by antibody cross-competition, using new information to inform immunogenicity studies (Aim 3). We will also map the specificities of cross-neutralizing and autologous neutralizing plasmas generated during natural HIV-1 infection. In Specific Aim 3, we will evaluate various VLP presentations in a program of rabbit and macaque immunizations. The theme of our approach is immunomodulation. VLPs will be presented in ways to try to enhance Ab trimer responses. Primarily, we will try to favor anti-trimer (neutralizing) responses by attempting to eliminate Ab responses to nonfunctional Env and cellular proteins embedded in membranes. We will examine the effect of precomplexing VLPs with antibodies. We will modulate antigen presentation in various ways. We will also examine the propensities of different Envs to elicit nAbs, including those cloned from patients who developed cross-neutralizing responses.